Community Health Evaluation clinics general practice in ...eprints.kingston.ac.uk/17255/1/Bowling-A-17255.pdf · Evaluation ofspecialists' outreach clinics in generalpractice year

52ournal of Epidemiology and Community Health 1997;51:52-61

Evaluation of specialists' outreach clinics ingeneral practice in England: process andacceptability to patients, specialists, and generalpractitioners

Ann Bowling, Katia Stramer, Edward Dickinson, Joy Windsor, Matthew Bond

Centre for HealthInformatics andMultiprofessionalEducation (CHIME),University CollegeLondon MedicalSchool, 4th FloorArchway Wing,Whittington Hospital,Highgate Hill, LondonN19A BowlingK StramerJ WindsorM Bond

Research Unit, RoyalCollege of Physicians,LondonE Dickinson

Correspondence to:Dr A Bowling.

Accepted for publicationMay 1996

AbstractObjectives - The wider study aimed to

evaluate specialists' outreach clinics in re-

lation to their costs, processes, and effect-iveness, including patients' and profes-sionals' attitudes. The data on processes

and attitudes are presented here.Design - Selfadministered questionnaireswere drawn up for patients, their generalpractitioners (GPs) and specialists, andmanagers in the practice. Information wassought from hospital trusts. The studyformed a pilot phase prior to a wider evalu-ation.Setting - Nine outreach clinics in generalpractices in England, each with a hospitaloutpatient department as a control clinicwere studied.Subjects - The specialties included were

ear, nose, and throat surgery; rheum-atology; and gynaecology. The subjectswere the patients who attended either theoutreach clinics or hospital outpatientsclinics during the study period, the out-reach patients' GPs, the outreach patients'and outpatients' specialists, the managersin the practices, and the NHS trusts whichemployed the specialists.Main outcome measures - Process itemsincluded waiting lists, waiting times inclinics, number of follow up visits, in-vestigations and procedures performed,treatment, health status, patients' andspecialists' travelling times, and patients'and doctors' attitudes to, and satisfactionwith, the clinic.Results - There was no difference in thehealth status of patients in relation to theclinic site (ie, outreach and hospital out-patients' clinics) at baseline, and all butone of the specialists said there were no

differences in casemix between their out-reach and outpatients' clinics. Patientspreferred, and were more satisfied with,care in specialists' outreach clinics in gen-eral practice, in comparison with out-

patients' clinics. The outreach clinics wererated as more convenient than outpatients'clinics in relation to journey times; thoseoutreach patients in work lost less timeaway from work than outpatients' clinicpatients due to the clinic attendance.Length of time on the waiting list was

significantly reduced for gynaecologypatients; waiting times in clinics were

lower for outreach patients than out-patients across all specialties. In addition,outreach patients were more likely tobe first rather than follow up attenders;rheumatology outreach patients weremore likely than hospital outpatients toreceive therapy. GPs' referrals to hospitaloutpatients' clinics were greatly reducedby the availability ofoutreach clinics. Bothspecialists and GPs saw the main ad-vantages of outreach clinics in relation tothe greater convenience and better accessto care for patients. Few of the specialistsand GPs in the outreach practices heldformal training and education sessions inthe outreach clinic, although over half ofthe GPs felt that their skills/expertise hadbroadened as a result of the outreachclinic.Conclusions - The processes ofcare (wait-ing times, patient satisfaction, conven-ience to patients, follow up attendances)were better in outreach than in out-patients' clinics. However, waiting listswere only significantly reduced for gyn-aecology patients, despite both GPs andconsultants reporting reduced waiting listsfor patients as one of the main advantagesof outreach. Whether these improvementsmerit the increased cost to the specialists(in terms of their increased travellingtimes and time spent away from their hos-pital base) and whether the developmentof what is, in effect, two standards of carebetween practices with and without out-reach can be stemmed and the standard ofcare raised in all practices (eg, by sharingoutreach clinics between GPs in an area)remain the subject of debate. As the datawere based on the pilot study, the resultsshould be viewed with some caution, al-though statistical power was adequate forcomparisons of sites if not specialties.

(J7 Epidemiol Community Health 1997;51:52-61)

Considerable interest has been expressed inestablishing specialist clinics in general prac-titioners' (GPs) surgeries, health centres, andcommunity clinics (known as "outreach") asone method of shifting the balance of care fromthe secondary to the primary care sector. Thereare approximately 40 million hospital out-patient attendances in acute specialties each

52

Evaluation of specialists' outreach clinics in general practice

year alone in the UK, costing 1200 millionin 1990.' Clinics often have long waiting listsbecause ofthe number of re-attendances, manyof which may be inappropriate.23 While theremay often be good reasons for the provisionof continuing care in hospital clinics, otherreasons include consultants' lack of confidencein GPs' ability to manage their patients, poor

inter-professional communication, and the lowpatient discharge rate by junior hospital doc-tors."7 One of the main conclusions from Cart-wright and Windsor's national outpatientsurvey3 was that the balance of outpatient at-tendance was inappropriate - there were toomany continuing attendances over long periodsand too few single or short term attendanceswith referral back to GPs.The concentration on the shift ofappropriate

health services from secondary to primary care,which is being emphasised in London in par-

ticular,>'2 has drawn attention to the possibilitythat much of the outpatient work could takeplace in GP surgeries rather than in hospitals.It is thus envisaged that hospital outpatientclinics will consist of those aspects of secondarycare which need specialised technical activityor the use of expensive equipment.9 If this issuccessful, referrals to hospitals should show a

downward trend, with more patients receivingfollow up care from their own GPs or fromspecialists consulting in primary care centres(outreach clinics). Specialist clinics in primarycare settings are not new, but there has beenlittle thorough evaluation of them. Given thepredicted growth in the numbers ofthese clinicsin the near future, an evaluation is essential.The establishment ofspecialist outreach clin-

ics in general practice has been given impetusby fundholding GPs who have initiated themin a wide range of specialties.'' Bailey, Blackand Wilkin's survey of 50 provider units inEngland'3 identified 96 outreach clinics insurgical specialties, medical specialties, andpsychiatry, and they concluded that their de-velopment is likely to increase. They reportedthat there was little evidence from their de-scriptive survey that GPs attended the outreachclinics or had any face to face communicationwith the specialists. While specialists and GPsreported that the greatest benefit for patientsincluded ease of access and shorter waitingtimes for outpatients appointments, some spe-

cialists reported problems of having to makerepeat appointments for patients who neededhospital based investigations, consultants' trav-elling times to outreach clinics, and more re-

stricted time available for training juniorhospital doctors. A quarter of the specialistssaw the clinics as a means of attracting incomeand referrals to their hospitals. Just a fifth ofthe outreach clinics in fundholding practiceswere open to referral from other practices,compared with almost two thirds of the clinicsheld in non-fundholding practices. The authorsconcluded that there is still no firm evidenceabout whether these clinics make an importantcontribution to overcoming the barriers be-tween primary and secondary care, or abouttheir cost effectiveness. Despite this, they re-

ported that both specialists and GPs had plans

for extensive further development of specialistoutreach clinics.There is much to learn from the precedent of

psychiatry.'5 Simply shifting outpatient sessionsto primary care settings does not, by itself,enable GPs and consultants to influence eachother or facilitate joint decision making. Thisis because the most common model is the"shifted outpatient" model in which the spe-cialist conducts a normal outpatient clinic ingeneral practice premises, often at a time whenthe GP is not on the premises so that contactis therefore infrequent.'6 An improvement onthis is the consultation-attachment in whichthe specialist attends a primary care meetingto discuss the management of several difficultpatients with primary care staff, after which thespecialist sees several patients, sometimes withthe GP.'7 With this model the GP continues toprovide treatment for the patients, but benefitsfrom joint management plans and specialistadvice on patients whom he or she does notwish to refer. This method of care has thepotential of being more expensive. Apart fromthe issue of cost effectiveness, the issue of how,or whether, to meet the demand for outreachclinics from an increasing number of practiceshas not been debated. There are ethical con-cerns about the provision of outreach in se-lected practices (usually fundholders), leadingto a "two tier service".The results presented here are based on a

pilot study of the processes (including at-titudes), costs, and effectiveness of specialistoutreach clinics in general practice in com-parison with outpatient controls. This paperfocuses on the process and attitude data.

Aims and objectivesThe study aimed to describe the processes(including patients' and doctors' attitudes) ofoutreach clinics and to evaluate the costs andeffectiveness of specialist outreach clinics ingeneral practice. In relation to the processes,the hypotheses of the study were that specialistoutreach clinics will:1. Improve access for patients to specialistcare, reduce waiting times for appointments,thereby having a positive impact on short termoutcomes, and increase patient satisfaction;2. Improve communication between spe-cialists and general practitioners, and have edu-cational benefits for general practitioners,thereby also increasing professional sat-isfaction;3. Reduce GPs' referrals to hospital outpatientdepartments;4. Reduce the number of follow-up visits tothe specialist, thereby enhancing the shift ofcare and workload from the secondary to theprimary care sector.

MeasuresThe measures used were based on well testedquestionnaires and items -for measuring pro-cesses, attitudes, and satisfaction, includingDavies and Ware's consumer satisfaction ques-tionnaire (based on questions tested for the

53

Bowling, Stramer, Dickinson, Windsor, Bond

RAND medical outcomes study in the USA).`8Health status was measured using the RANDbrief impact and outcome batteries'9 and the12 items from the RAND version of the shortform 36 health status/health related quality oflife questionnaire that make up the recentlydeveloped health status questionnaire-12.20Other disease specific items of relevance andsociodemographic data were included in thequestionnaires, including a list ofhealth relatedquality of life items generated by the public(areas of life most affected by longstandingillness).2' A simple visual analogue scale wasused for the specialists' and GPs' ratings ofseverity of the patients' condition, based on thedefinitions used in the Duke severity of illnessscale.22 The questionnaires were further pre-tested before the pilot study for acceptability,comprehension, and content validity on 60outpatients and volunteers from patients'groups, and on medical staff in the Departmentof General Practice at St Bartholomew's Hos-pital Medical College (where the research teamwas initially based) before being used in thestudy reported here. The process, attitude, andcost questions for patients, specialists, and GPswere developed with colleagues at the PrimaryCare Research and Development Centre, Uni-versity ofManchester with whom these authorsare collaborating in relation to the main evalu-ation of outreach clinics across England indifferent specialties (cost data are not reportedhere as the analysis is continuing). Base num-bers to questions may vary due to some itemnon-response.

MethodsThis study presents the results on processesand attitudes from nine outreach clinics inEngland, each with same speciality outpatients'clinics as controls (they were the same spe-cialists' outreach and outpatients' clinics, ex-cept in two cases (one in gynaecology and onein rheumatology) where the specialists were infull time private practice and so local NHSoutpatients were used as the controls). Theadvantage of the same specialists' outpatients'clinic acting as the control clinic is to reducevariation (eg in style of clinical practice), al-though this reduction will not be achieved inthe case of two of the clinics. The advantage,however, of including a small number of un-paired outpatients' clinics as controls is thatthe clinics included in the study will reflect thevariation in clinic providers that exists in reality(ie some run by NHS specialists and a few byprivate specialists), facilitating analyses of themost appropriate model of providing outreach.In the main study, the unpaired clinics can beanalysed separately in order to assess the extentof any variation effects.The specialties included were rheumatology

(three outreach clinics), ear, nose and throat(ENT) (three outreach clinics), and gyn-aecology (three outreach clinics). The outreachclinics included in the study were selected torepresent a wide geographical spread of regionsin England. They were selected after identifyingthe location of outreach clinics in England from

a screen of family health services, acute trustexecutives and selected specialists throughtheir specialist bodies. The data from the studyreported here formed the pilot phase of thestudy. The questionnaires were unchanged forthe main study, which is currently underway.

In view of the predicted growth in the num-bers of specialist outreach clinics (unpublisheddata from a postal screen of specialists andNHS trusts as part of this study), and the lackof any evaluative data about such clinics, thedissemination of the pilot results was felt to beworthwhile. It can sometimes be an error ofjudgement to disseminate pilot findings, in theevent that main study findings differ. However,while the main study will be the largest andmost comprehensive evaluation of outreachclinics (although it will not be completed fortwo years), the analyses from the pilot studypresented here also form the largest study todate. The pilot study involved the collection ofdata from nine outreach clinics with outpatientcontrols across England. The study included146 outreach patients and 148 outpatients.This provides sufficient statistical power forcomparisons between sites (outreach and out-patients), although statistical power will beweaker for between specialty comparisons. Forthe main study we require 1000 outreachpatients and 1000 outpatients in more than 30paired clinics (outreach-outpatients) to achievestatistical power with the clinic and the patientas the unit of analysis and to permit betweenspecialty comparisons. This is ongoing. We willbe undertaking multilevel analysis for the mainstudy, and we will be able to increase thenumber of stratified analyses to control forintervening variables.

Pilot study analyses are necessarily crude incomparison with main study analyses becausethe statistical power is lacking for fine strat-ification within the analyses and because theinvestigator wishes to limit the amount of stat-istical testing carried out in order to reducethe potential for statistical significance beingobtained by chance. Thus, as the data arebased on a pilot study, conclusions can onlybe tentative and must be viewed with caution.

In each participating outreach clinic andmatched outpatients' clinic, all attendingpatients were approached in the waiting roomand invited to take part. They were given a selfcompletion questionnaire to take home andreturn to the research team. Specialists andGPs completed clinical sheets for the patientsas well as process and attitude questionnairesabout the outreach clinic. The practice man-agers and the NHS trusts provided process andcost data.

RESPONSE RATESIn relation to the nine outreach clinics, andtheir matched outpatients' clinics, each of thenine practice managers returned their ques-tionnaires about the costs and processes of theoutreach clinic. All of the trusts in each studyarea provided information about outpatientcosts. Each ofthe nine specialists returned theirattitude questionnaires. Forty four (73%) of

54


the 60 GPs in the study practices with outreachclinics returned their attitude questionnaires.Nineteen visits were made to the nine differ-

ent outreach clinics, and one visit each to theircorresponding outpatients' clinics (controls)were made over a three month period in orderto recruit the samples. More visits were madeto recruit patients in the outreach clinics be-cause ofthe smaller numbers ofpatients bookedin, in comparison with outpatients' clinics. Onehundred and forty six (83%) of the 176 out-reach clinic patients attending (all attenderswere approached) returned their questionnaires(ENT: 80%, 65 out of 81; gynaecology: 83%,44 out of 53; rheumatology: 88%, 37 out of42), as did 148 (71 %) out ofthe 208 outpatientsattending (all attenders were approached)(ENT: 77%, 66 out of 86; gynaecology: 63%,43 out of 68; rheumatology: 72%, 39 out of54).

In relation to the two sided individualpatient's clinical sheets completed by GPs (foroutreach clinic patients) and by specialists (foroutreach clinic and outpatients' clinic patients),the response rates were 58% (102 out of 176returned) for the GPs (outreach clinic patients- and a further 15 were lost in the post); 96%(169 out of 176 attending) for the specialistsin relation to outreach clinic patients, and 82%(170 out of 208 attending) for the specialistsin relation to the outpatient controls (as theycompleted clinical sheets in cases where thepatient failed to return their questionnaire, thenumbers of clinical sheets exceeds the totalpatient response rates). These response rateswere considered to be excellent in view of thebusy work schedules of clinicians and dem-onstrate the level of interest in, and com-mitment to, the study.

ResultsINFORMATION FROM THE MANAGERS IN THEPRACTICE, THE GPS AND THE SPECIALISTSAll the participating practices were fundholdersor multi-fundholders, and all had outreach clin-ics in other specialties. All but one of theoutreach clinics had been established betweenone and four years ago (one rheumatologyoutreach clinic had been operating for less thana year). Five of the outreach clinics were heldmonthly, one gynaecology clinic was held everyfortnight, two ENT clinics were held every sixweeks, and one rheumatology clinic was held"as required". Neighbouring practices could,in theory, refer their patients to two of theoutreach clinics; in practice this was rare. Apartfrom two practices which paid the specialist orhospital trust a fee per patient booked (1C35-£40 per patient), the remaining practices paida set clinic fee, regardless of the number ofpatients booked (£230-£540 per clinic). Thepractice also had to bear the costs of any in-vestigations or procedures performed in out-reach that require additional facilities (eg,routine tests requiring laboratory analysis, re-ferral to hospital for further investigations orprocedures). The average trust charges for out-patient care in the study districts (in the studyspecialties) was £69 for a new referral (range

£48-£89, except for one trust which absorbedthe outpatient cost within the inpatient fee)and £42 for a follow up consultation (range£26-£64, except for one trust which absorbedthe outpatient cost within the inpatient fee).The charges include basic investigations. Thecostings of the clinics have yet to be com-prehensively analysed, thus the costs reportedabove must be viewed with caution as they arecrude and form a partial reflection only of truecosts.

All of the specialists providing outreach heldconsultant status. Two of the specialists pro-viding outreach services were in full time privatepractice, and the remainder held NHS ap-pointments. Six of the eight specialists whoreplied said the outreach clinic was set up at theGPs' suggestion, one said it was a fundholdingconsortium's suggestion, and one said it wasthe trust's suggestion. The specialists' statedreasons for setting up the clinics were mainlypatient oriented - all said it was to reducewaiting list times for a specialist's opinion and/or to improve accessibility for patients; a third(3) also said they set up the outreach clinic inorder to secure GP-fundholder contracts forthe hospital. The most frequent reasons givenby the GPs for setting up the outreach clinicwere: to improve accessibility/convenience forpatients (94%, 33); to get priority access to aparticular consultant (83%, 29); to improvecommunication between specialists and GPs(74%, 26); to reduce waiting times for spe-cialists' opinions (71%, 25); to broaden GPs'skills (57%, 20); and to improve GPs' jobsatisfaction (43%, 15).Three of the specialists travelled 20-23 miles

to the outreach clinic, two travelled 35-40miles, and the remainder travelled less than 10miles (their return journeys were similar). Themanagers reported that each outreach cliniclasted between 2 and 3.5 hours, although threeofthe specialists reported that the clinic actuallylasted for 4 hours. In relation to the totalamount of time devoted to the outreach clinicon the day the clinic was held, four of thespecialists reported devoting between 5 and 6hours to it and the remainder devoted 2.5 to5 hours to it. Three specialists reported thattheir outreach clinic was conducted in normalNHS time; three said it was done in privatetime; and the remainder said it was done inextra NHS sessions. Five of the specialists re-ported that they held outreach clinics in otherpractices, and one was planning these.

In relation to outreach patients requiringfurther tests/investigations in hospital, five spe-cialists said they gave them the next availableappointment (thus, in effect, giving them a"fast track"), and the remainder said that theyput the patients on the waiting list and treatedthem as new referrals.

REFERRALS TO OUTREACHThe managers in the practices provided in-formation on the average number of patientsbooked into the outreach clinics: this was 15.66(range 10-25) (ENT: 19.66, range 12-25;rheumatology: 14.0, range 10-16; gynaecology:

55


Table 1 GPs' view of the main advantages and disadvantages of outreach clinics

ENT Gynaecology Rheumatology Total %(no) (no) (no) (no)

Advantages:Reduces waiting times for patients to get appointments 7 13 7Improves accessibility/convenience for patients 7 3 15 77 (27)Improves communication between GP and specialist 6 11 10 71 (25)Broadens GPs' skills 2 7 9 77 (27)Broadens specialists' skills 1 5 9 51 (18)Improves GPs' (own) job satisfaction 5 8 12 42 (15)Fewer non-attenders in comparison 4 10 2 63 (22)Promotes good will with specialist 4 8 2 40 (14)Cheaper service than outpatients' clinics 4 10 3 49 (17)

Disadvantages:Having to repeat appointments for patients who need tests at hospital 3 1 3 20 (7)Lack of equipment in GPs' surgery - 3 - 9 (3)Lack of surgery space/rooms 2 2 1 14 (5)Increase in GPs' administrative costs/time 4 5 8 49 (17)Reduces specialists' time in hospital 2 3 9 40 (14)No disadvantages 1 3 4 23 (8)No of respondents 7 13 15 35

ENT= ear, nose, and throat.

13.33, range 10-15). All but one of the GPssaid the patients they referred to the outreachclinic were patients who they would otherwisehave referred to the hospital outpatients' de-partment, rather than have managed themthemselves (one said that the outreach clinicpatients were a combination of those theywould have managed themselves or referred tohospital). All except one (in ENT) of the NHSspecialists said that the casemix of their out-reach and outpatients clinics was similar (theone in ENT said that he saw less acute patientsin outreach clinics). The practice managers pro-vided information on the number of patientsreferred to (all) hospital outpatients' departments(in the same specialty as the outreach clinic)during the six months before and after the out-reach clinic had been set up. The volume ofreferrals had decreased after the outreach clinicstarted. The average number of outpatient re-ferrals per practice in the six months before theoutreach clinics were initiated was 82.0 (range62-246) and in the six months after the clinichad started the average referral rate to outpatientswas 9.0 (range 0-17) (t test not performed asnumber of clinics in sample was small).

COLLABORATION AND CONTACT BETWEENPROFESSIONALSSix of the specialists said the GP decided whichpatients were to be seen in outreach clinicsand the remainder said it was a joint decisionbetween GP and specialist. Only one of thespecialists reported having (joint) criteria/guidelines for the type of patient to be seen inoutreach (ie, those with non-acute conditions).Four of the GPs said they decided jointly withthe specialist who should be discharged fromthe outreach clinic and the remainder said thespecialist alone decided. When asked who hadoverall responsibility for an outreach clinicpatient, 67% (22) of the GPs said that theyretained responsibility and the remainder saidthe specialist held responsibility.

In six of the nine outreach clinics the spe-cialist was accompanied by other staff. Onegynaecologist reported seeing patients with thepractice nurse (eg, as chaperone); one gynae-cologist reported that his private nurse andprivate secretary accompanied him to the out-

reach clinic; in two of the rheumatology out-reach clinics the specialist was accompanied bytwo NHS hospital nurses (one did blood testsand the other helped patients to dress/undress);two hospital based audiologists tested the hear-ing of patients before the consultation with thespecialists in two ENT outreach clinics.Two of the specialists reported periodically

holding educational and training sessions"teach and treat" with the GPs in the outreachclinic. Otherwise, none of the specialists hadplanned meetings with the GPs (com-munications were by letter, fax, and telephone).

GPS ATTITUDES TO OUTREACHFifty three per cent (18) of the GPs felt thattheir skills/expertise had been broadened as aresult of the outreach clinic, 35% (12) felt theyhad not, and 12% (4) were uncertain. Fourteenper cent of the GPs (5) were planning otheroutreach clinics.GPs were asked about the advantages and

disadvantages of the outreach clinic, and theirresponses are shown in table 1. The mostcommonly stated advantages (by over half)were the reduced waiting times for patientsto get appointments; improved accessibility/convenience for patients; fewer non-attendersthan in outpatients; improved job satisfactionfor GPs; and improved communication be-tween GPs and specialists.

Table 1 also shows the GPs' perceived dis-advantages of outreach clinics. The largest cat-egory related to the increase in GPs'administration costs/time, this was followedby reduced time in hospital (NHS) for thespecialist and having to make repeat ap-pointments for patients who need tests/in-vestigations in hospital. Twenty three per cent(8) ofthe GPs said there were no disadvantagesof outreach clinics. All but one (in rheum-atology) of the GPs said that they believed thatthe outreach clinic was worthwhile (96% (23)).

SPECIALISTS ATTITUDES TO OUTREACHThe specialists perceived fewer advantages ofoutreach than the GPs reported. The mostcommonly reported advantages were reducedwaiting times for patients to get appointments

56


Table 2 Time on waiting list to see specialist and waiting times within the clinic in relation to clinic site

Outreach Outpatient

ENT Gynaecology Rheumatology Total ENT Gynaecology Rheumatology Total% (no) % (no) % (no) % (no) % (no) % (no) % (no) % (no)

Time on waiting list tosee specialist:<3wk 4 (2) 53 (21) 15 (5) 21 (27) 21 (10) 15 (5) 5 (2) 15 (17)3<5wk 17 (10) 33 (13) 26 (8) 24 (31) 21 (10) 18 (6) 36 (10) 24 (26)5<7wk 24 (14) 10 (4) 28 (9) 21 (28) 10 (5) 32 (11) 11 (3) 18 (19)7<9wk 22 (13) - 12 (4) 13 (17) 8 (4) 6 (2) 11 (3) 8 (9)9<36wk 35 (20) 5 (2) 18 (6) 21 (27) 40 (19) 29 (10) 36 (10) 35 (39)No of respondents 58 40 32 130 48 34 28 110Waiting times within theclinic:No wait 20 (12) 45 (17) 43 (14) 33 (43) 2 (1) 32 (12) 9 (3) 12 (16)1-lOmin 23 (14) 34 (13) 36 (12) 30 (39) 8 (4) 19 (7) 9 (3) 11 (14)11-20min 20 (12) 10 (4) 12 (94) 15 (20) 13 (7) 19 (7) 3 (1) 12 (15)21-40min 12 (7) 5 (2) 6 (2) 8 (11) 38 (21) 14 (5) 24 (8) 27 (34)41-60min 17 (10) 3 (1) 3 (1) 9 (12) 15 (8) 11 (4) 20 (7) 16 (19)61-150min 8 (5) 3 (1) - 5 (6) 24 (13) 5 (2) 35 (12) 22 (27)No of respondents 60 38 33 131 54 37 34 125

ENT = ear, nose, and throat.

(8); improved communication between GPsand specialists (6); and promotes goodwill withGPs (6). The most commonly reported dis-advantages were the travelling times for thespecialist (6), followed by reduced specialists'time in hospital (NHS) (5), as well as ontraining junior doctors (4), and having to makerepeat appointments for patients who requiretests on the hospital site (4). All but three ofthe nine specialists (two rheumatologists andone gynaecologist) said that they thought theoutreach clinic was "worthwhile".

THE PATIENTS: MEDICAL CONDITION ANDPREVIOUS CLINIC ATTENDANCESThere were no differences in relation to site(outreach or outpatients' clinic) and the typeof medical condition (diagnosis was coded ac-cording to the International Classification ofDiseases, 10th version), patients' reports ofimpact on quality of life, self assessed physicaland mental health status, or the length of timepatients had suffered from their condition. Forexample, 22% (16) of employed outreachpatients and 17% (11) of outpatients had takenthree weeks or more off work in the past sixmonths because of ill health. Forty eight percent (66) of outreach patients and 52% (64)of outpatients said they had "accomplished lessin work/other daily activities due to physicalhealth" in the past four weeks. Twenty sevenper cent (37) of outreach patients and 29%(36) of outpatients said they had no pain inthe past four weeks. Nine per cent of outreachpatients (12) and 9% (11) of outpatients saidthat over the past four weeks they had felt "fullof life" none of the time. These results confirmall but one of the specialists' reports of nodifferences in casemix in relation to site (seeearlier).More (65% (89)) of the outreach patients

than outpatients (34% (45)) said the sampledconsultation was the first time they had at-tended the ("this") specialist clinic for theircondition (X2: 25.63; 1 df; p<O.OOO1). Thiswas only significant, however, for ENTpatients: 72% (44) of ENT outreach patientsand 22% (13) of ENT outpatients said this

was their first attendance (x2: 29.44; 1 df;p<0.OO 1).Of all the follow up patients attending clinics

that had been in operation for a year or more,23% (11) of the outreach clinic patients and40% (35) of the outpatients had first attendedthe clinic more than a year ago (X2: 4.13; 1 df;p<O.05).

TIME ON THE WAITING LIST AND WAITINGTIMES IN CLINICThe differences in waiting times for the totalsamples in relation to site were not statisticallysignificant (neither at fewer than three weeksor at nine or more weeks). However, there weredifferences in relation to specialty. Table 2shows that 53% (21) of outreach patients ingynaecology waited less than three weeks tosee the specialist in comparison with 15% (5)of gynaecology outpatients (X2: 11.52; 1 df;p<O.OO1). Differences withinENT and rheum-atology in relation to site were not statisticallysignificant. With regard to ENT, outreachpatients appeared to be less likely than out-patients to be seen within three weeks (notstatistically significant). This partly reflectedthe lesser frequency with which the outreachclinics were held (two of the three ENT out-reach clinics were held every six weeks, incomparison with monthly for most of the otheroutreach clinics). Two of the rheumatologyoutreach clinics were organised by the hospital(where appointments were made, rather thanby practice staff) and this may explain the lackof difference between sites (ie, these practiceswere not permitted the flexibility of havinglarger clinics when needed).There were differences between sites, and

within specialty groups between sites, in thelength of time, after the appointment time, thatpatients had to wait at the clinic before seeingthe specialist (see table 2). More ofthe outreach(33%, 43) than outpatients' clinic patients(12%, 16) waited for 10 minutes or less (X2:8.10; 1 df; p<0.0 1), while the outpatients weremore likely to wait for one hour or more (22%,27) in comparison with outreach patients (5%,6) (X2: 14.54; 1 df; p<0.001).

57


Table 3 Satisfaction with the visit to the specialist clinic (row %o): ear, nose, and throat, gynaecology, and rheumatolog combined

Outreach patients (n= 132-138) Outpatients (n= 126-130)

Excellent Very good Good Fair Poor Excellent Very good Good Fair Poor% (no) % (no) % (no) % (no) % (no) % (no) % (no) % (no) % (no) % (no)

Length of time to get appointment with 27 (37) 27 (36) 26 (35) 13 (18) 7 (10) 10 (13) 22 (29) 34 (44) 25 (33) 9 (11)specialist*

Convenience of location of clinic* 54 (73) 30 (40) 9 (13) 7 (9) - 14 (18) 18 (24) 36 (47) 22 (28) 10 (13)Getting through to clinic by phonet 26 (11) 26 (11) 29 (12) 12 (5) 7 (3) 16 (9) 35 (20) 32 (18) 15 (9) 2 (1)Length of time waiting at clinic to see 36 (48) 23 (31) 17 (23) 16 (21) 8 (12) 10 (12) 17 (22) 24 (30) 26 (33) 23 (29)

specialist*Time spent with the specialist* 27 (36) 27 (36) 30 (40) 11 (14) 5 (7) 10 (13) 24 (31) 40 (50) 21 (27) 5 (7)Explanation of what was done* 30 (40) 28 (37) 27 (35) 11 (15) 4 (5) 20 (25) 33 (42) 27 (34) 11 (14) 9 (11)Thoroughness, carefulness, competence of 40 (54) 27 (37) 24 (32) 7 (10) 2 (3) 33 (43) 31 (40) 21 (27) 12 (15) 3 (4)

specialistPersonal manner (courtesy, respect, 45 (61) 30 (41) 15 (20) 7 (9) 3 (4) 40 (51) 32 (41) 22 (28) 3 (4) 3 (5)

sensitivity, friendliness) of specialistEase of making or changing appointmentt 31 (9) 34 (10) 14 (4) 14 (4) 7 (2) 15 (7) 38 (18) 27 (13) 14 (7) 6 (3)Convenience of appointment day/time* 10 (10) 41 (42) 41 (43) 7 (7) 1 (1) 5 (7) 38 (48) 40 (51) 14 (18) 3 (4)Waiting area and facilities* 30 (42) 33 (46) 27 (37) 7 (9) 2 (4) 5 (6) 31 (40) 40 (52) 18 (24) 6 (8)Attention given to what you had to say* 35 (48) 34 (47) 21 (29) 7 (10) 3 (4) 17 (22) 37 (47) 31 (40) 11 (14) 4 (5)Personal manner (courtesy, respect, 35 (47) 34 (46) 20 (28) 10 (13) 1 (2) 22 (28) 41 (52) 28 (36) 9 (12) -

sensitivity, friendliness) of reception staffAdvocacy/interpreter facilities§ - (4) - - - - - (1) - (3) - (1) - -

Overall visit 31 (42) 40 (54) 20 (27) 8 (10) 1 (1) 17 (22) 36 (46) 28 (35) 16 (20) 3 (4)

* Statistically significant at least at p<0.01 (%2).tTelephone base: 42 outreach, 100 outpatients.t Changing appointment base: 27 outreach and 48 outpatients.§ Advocacy base: 4 outreach and 5 outpatients.

OUTCOME OF THE CONSULTATION

The outreach patients were more likely than

outpatients to be first attenders, and there weredifferences in the percentages given a follow

up appointment after the sampled clinic visit:37% (50) of outreach patients and 50% (66) of

outpatients were given a follow up appointment(X2:5.04 ldf; p<0.05).More of the outreach patients' GPs than the

outpatients' GPs were reported to have sent

the specialist the results of tests/investigationswhen the patient was referred (26% (40) versus

13% (19) - X2: 8.29; 1 df; p<0.01), althoughdifferences within specialty were only apparentfor ENT (26% (18) of outreach patients' GPs

sent the specialist the results of tests, in com-

parison with none of the outpatients' GPs).Outreach patients were also less likely to have

any tests requested by the specialist than out-

patients: 30% (32) had one or more tests re-

quested by the specialist, in comparison with

57% (61) of outpatients (X2: 16.11; 1 df;p<0.001). These differences were evident

within each specialty except in ENT, where

(excluding routine audiology testing prior to

the consultation) outreach patients were more

likely to have tests than outpatients (ENT: 27%(11):4% (1) - x2 not performed as there were

fewer than 5 expected cases in a cell). In gyn-aecology, 9% (3) of the outreach patients and

39% (19) of the outpatients had tests (X2 not

performed as there were fewer than 5 expectedcases in a cell), and in rheumatology, 57% (19)of the outreach patients and 81% (26) of the

outpatients had tests (x2: 4.28; 1 df; p<0.05).Specialists were asked if they had prescribed

or suggested any treatment for the patients.They reported they had done so for 76% (118)of the outreach patients and for slightly fewer

(67% (108)) of the outpatients (X2: 4.39; 1

df; p<0.05). The difference was significant for

rheumatology patients. Among rheumatologypatients, outreach patients were more likelyto be referred for therapy (37% (13)) than

outpatients (14% (5) - x2: 5.36; 1 df; p<0.05).The difference was also apparent for both ENT

and gynaecology patients, but did not achievestatistical significance. For example, outreachENT patients were slightly more likely to begiven some treatment (usually medication or

surgery) than ENT outpatients (75% (53) ver-

sus 61% (37); not significant). Among gyn-aecology patients, outreach patients were more

likely to be referred for surgery (56% (28))than outpatients (38% (24); not significant).There were no differences between sites, or

specialties, in numbers of types of medicationsprescribed; nor were there any significantdifferences in numbers of "over the counter"medications purchased.

PATIENTS' PREFERENCES AND SATISFACTIONAll patients were asked where they preferred tosee the specialist - at the GP's surgery, at thehospital, or whether they had no preference.Altogether 73% (101) of outreach patients (72-73% within each specialty) said they preferredthe GP's surgery, 1% (2) said they would havepreferred the hospital, and 26% (36) reportedno preference. In comparison, 44% (63) of theoutpatients said they would have preferred to

have been seen in the GP's surgery, 22% (31)said they preferred the hospital, and 34% (49)said they had no preference (x2, preference forGP's surgery: 23.70; 1 df; p<0.0001).

Outreach clinic patients were less likely thanoutpatients clinic patients to say they wouldlike to see something improved in the clinic(11% (14) versus 22% (26) - x2: 4.32; 1 df;p<0.05).The results for the patients' satisfaction items

are presented in relation to site only in table3. Outreach patients were more satisfied thanoutpatients with the clinic visit in relation to

the length of time to get an appointment withthe specialist, the convenience of the locationof the clinic, the length of time waiting at theclinic to see the specialist, the amount of timespent with the specialist, the convenience ofthe appointment day/time, the waiting areas

and facilities, and attention given to what the

58


Table 4 Distance from outreach clinic and journey times to and from the clinic in relation to specialty and site

Outreach Outpatient

ENT Gynaecology Rheumatology Total ENT Gynaecology Rheumatology Total% (no) % (no) % (no) % (no) % (no) % (no) % (no) % (no)

Journey distance:To clinic:<3 miles 68 (39) 60 (24) 54 (19) 62 (82) 15 (8) 30 (11) 53 (17) 29 (36)3<5 miles 12 (7) 24 (10) 31 (11) 21 (12) 22 (12) 30 (11) 28 (9) 26 (32)5<7 miles 3 (2) 3 (1) 6 (2) 4 (5) 13 (7) 8 (3) 13 (4) 12 (14)7<15 miles 12 (6) 10 (4) 6 (2) 9 (12) 24 (13) 27 (10) 3 (1) 18 (24)15<20 miles 3 (2) - - 2 (2) 22 (12) 5 (2) 3 (1) 13 (15)20+ miles 2 (1) 3 (1) 3 (1) 2 (3) 4 (2) - - 2 (2)

From clinic:<3 miles 75 (43) 58 (23) 57 (20) 65 (86) 12 (16) 26 (10) 52 (16) 26 (32)3<5 miles 12 (7) 30 (12) 31 (11) 23 (30) 21 (11) 30 (11) 26 (8) 25 (30)5<7 miles 4 (2) - 6 (2) 3 (4) 15 (8) 14 (5) 16 (5) 15 (18)7<15 miles 5 (3) 12 (5) 3 (1) 7 (9) 27 (14) 25 (9) 3 (1) 18 (21)15<20 miles 2 (1) - - 1 (1) 15 (8) 5 (2) 3 (1) 12 (14)20+ miles 2 (1) - 3 (1) 1 (2) 10 (5) - - 4 (5)

No of respondents 57 40 35 132 52-54 37 31-32 120-123Journey times;To clinic:0-5min 32 (20) 22 (9) 30 (10) 29 (39) - 5 (2) 9 (3) 4 (5)6-10min 24 (14) 41 (17) 32 (11) 31 (42) 7 (4) 16 (6) 20 (7) 14 (17)11-20min 24 (14) 22 (9) 29 (10) 24 (33) 40 (22) 50 (19) 53 (18) 46 (59)21-60 min 20 (12) 15 (6) 9 (3) 16 (21) 42 (23) 29 (11) 18 (6) 31 (40)61-120min - - - - 11 (6) - - 5 (6)

From clinic:0-5min 37 (22) 12 (5) 24 (8) 26 (35) - 3 (1) 9 (3) 3 (4)6-10min 27 (16) 48 (19) 35 (12) 36 (47) 3 (2) 17 (6) 21 (7) 13 (15)11-20min 25 (15) 22 (19) 32 (11) 26 (35) 44 (23) 47 (17) 52 (17) 47 (57)21-60min 9 (5) 18 (7) 9 (3) 11 (15) 40 (20) 33 (12) 15 (5) 30 (37)61-120min 2 (1) - - 1 (1) 13 (7) - 3 (1) 7 (8)

No of respondents 59-60 40-41 34 133-135 52-55 36-38 33-34 121-127

patient had to say (significance levels rangedbetween p<0.01 to p<0.001) for each of theseitems (dichotomised as satisfied/other, with x2test). Outpatients were not significantly morelikely than outreach patients to express greatersatisfaction with any item.

PATIENTS JOURNEYS: DISTANCE, LENGTH OFTIME AND COSTS

Sixty two per cent (82) of outreach clinicpatients and 29% (36) of outpatients travelledless than three miles to the clinic (X2: 27.64; 1df; p<0.0001). Table 4 shows that outpatientshad to travel much further to the clinic.

Patients were also asked about their journeytimes to and from the clinic. Table 4 showsthat 60% (81) of the outreach patients hada shorter journey time of up 10 minutes, incomparison with 28% (22) of outpatients (X2:49.96; 1 df; p<0.0001). In contrast, 37% (45)of the outpatients had longer journey times ofover an hour in comparison with 12% (16) ofthe outreach patients (X2: 20.37; 1 df;p<O.O001).Outreach patients in each specialty were far

more likely than outpatients to rate the journeyas "very convenient" (71% (95) and 36% (48)respectively). Altogether 25% (34) of outreachpatients and 45% (59) of outpatients rated thejoumey as "fairly convenient" and 4% (5) ofoutreach patients and 20% (25) of outpatientsrated the journey as "fairly" or "very in-convenient" (%2: "very convenient": 31.90; 1df; p<0.001; "very/fairly inconvenient": 15.37;1 df; p<0.001). These differences were evidentwithin each specialty.These reduced journeys and journey times

have implications for patients' travelling costsand associated expenses (eg, arrangements forchild care, time off work). For example, ofthose who took time off work at all, 50% (20)

ofoutreach patients and 24% (9) ofoutpatients(fewer) took 1 hour or less off work, 25% (10)of outreach clinic patients and 32% (12) ofoutpatients took 2 hours off work, and theremainder (25% (10) of outreach and 44%(17) of outpatients) took more time than thisoffwork (X2: 1 hour or less: 5.78; 1 df; p<O.05).This information is being used within the cost-ing formula (cost data analysis ongoing).

DiscussionAs Bailey et al reported,'3 in relation to theirearlier survey ofmanagers and doctors involvedin outreach clinics, fundholding practices haveused their purchasing power to secure a betterservice for their patients, although this leads toa risk of developing two standards of carebetween fundholding and non-fundholdingpractices. In line with the findings of Bailey etal, the study reported here found that the mostcommon advantages of outreach perceived bydoctors were ease of access for patients andshorter waiting lists. However, waiting lists wereonly significantly reduced for gynaecologypatients, despite both GPs and consultants re-porting reduced waiting lists for patients as oneofthe main advantages ofoutreach regardless ofspecialty. This is one consequence of relativelyinfrequent clinics (eg, 4-6 weeks) in relationto the number of referrals. Also, in relation totwo of the rheumatology clinics, the hospitaltook responsibility for making patients' ap-pointments, and therefore GPs did not havethe flexibility of being able to book morepatients into the clinics when the need arose.Few of the specialists and GPs in the out-

reach practices held joint training and edu-cation sessions in the outreach clinic, althoughover half of the GPs felt that their skills/ex-pertise had broadened directly as a result ofthe outreach clinic. The casemix of patients in

59


outreach and outpatients clinics was similar inthe specialities studied (although one ENTspecialist reported that he saw fewer acutepatients in outreach). Moreover, the GPs wereapparently more involved in the care of theoutreach patients in comparison with out-patients (they were more likely to send thespecialist in outreach the results of tests, andthere were fewer specialist follow up visits inoutreach). Although only 58% of the patients'clinical sheets had been completed by GPs, theinformation was collected from specialists, whocompleted most of the clinical sheets, andtherefore the results were not affected by re-sponse bias.There was some indication that outreach

patients were more likely to be treated thanoutpatients, particularly in rheumatology(where they were more likely to be referred fortherapy). The interpretation ofthis is uncertain,particularly as were no differences in healthstatus or impact of the condition on quality oflife between sites. It is possible that fundholdershave easier access to therapeutic servicesthrough their purchasing powers (eg, one ofthe practices with a rheumatology outreachclinic also had a private physiotherapist for thepatients). All patients are being followed upat six months in order to assess short termoutcomes. This issue will be addressed in futureanalyses, along with the comparative costs ofoutreach and outpatient clinic care in the spe-cialties selected for study.

In comparison with outpatients' clinics, theprocesses of care were generally superior inoutreach - patients' convenience and sat-isfaction were increased while their financialand time costs were decreased. Apart fromgynaecology outreach patients, who had ashorter period on the waiting list than gyn-aecology outpatients, the relative infrequencywith which most outreach clinics were held(eg, usually 4-6 weeks) carried the cost ofno advantages over outpatients in relation tolength of time on the waiting list. Whether thereported improvements are judged to be worththe increased cost to the specialists in terms oftheir increased travelling times and time spentaway from their hospital base (with the con-sequent implications for hospital patients' care,other work, and teaching time) remains thesubject of debate. The data on true costs tothe practice, the specialists and trusts, and theshort term outcomes of patients have also yetto be analysed. The other contentious issue isthat of the rapid development of a two tierservice between practices with and withoutoutreach clinics (which may, in turn, reflectfundholding versus non-fundholding prac-tices). Currently, there is not enough specialisttime to provide outreach clinics in all generalpractices. The recent changes in specialist train-ing and accreditation are likely to increase thenumber of fully accredited specialists belowconsultant level, making an increase in thenumber of specialist outreach clinics in generalpractice likely. A few districts are attemptingto avoid any rivalry between practices by pro-viding outreach clinics in community hospitalsor large health centres for all GPs to share

within a local patch. The danger then may bethat if these "locality outreach clinics" becometoo large and divorced from personal contactwith the practices, they too may develop thesame disadvantages of the outpatients' clinicsthat they were designed to overcome (eg, longerwaiting lists, longer follow up periods).The broader findings of the study reported

here echo a previous survey on the criteria GPswanted to see in contracts for outpatients' carein which a high premium was attached to theimprovement of communications between spe-cialists and GPs and between specialists andpatients; consultants (or at least registrars) see-ing all new patients; reduction of clinic waitingtimes (<half an hour of appointment time),and the elimination of duplicated, unnecessaryinvestigations.23 By inviting specialists to runoutreach clinics in their surgeries, GPs canattempt to control many of these features anddirectly improve patient care. It was concludedin that survey (in 1991) that if trusts did notmeet GPs' demands for higher quality out-patient care, the consequence would be anincrease in GP fundholding, which would resultin the loss of a district based health needsperspective. The consequence has been moredrastic than this, with fundholding GPs pur-chasing consultant care within their practicesand threatening further the original concept ofan integrated and equitable NHS.

In all but two cases, the same consultants'outpatients' clinics were used as the controlclinic. The justification was that this pairingof clinics would reduce variation (eg, clinicalvariation). This was not possible in two casesin which the consultants entered private prac-tice full time. The method of attempting to usethe same consultant's outpatients' clinics wherepossible is methodologically sound as long asthe process of care in one type of clinic doesnot rebound on the other clinic. The method-ology of controlled trials makes the assumptionthat the experimental and control groups areindependent. However, the paired design ofthis study (except for two clinics) allows forinteraction between the groups (outreach andoutpatient clinics) as both are the responsibilityof the same consultant. It is foreseeable thatthe casemix in outpatients' clinics might altertowards more severe or more complex cases ifall minor cases are dealt with in outreach clinics.However, as the number of patients seen inoutreach clinics still represents only a tiny frac-tion of the numbers seen in outpatients' clinics,it is unlikely that there is any significant inter-action between the two clinics.

Finally, as the data presented here are partof a pilot study, conclusions can only be tent-ative and must be viewed with caution. Theywere reported, before the completion of themain study, because of the relative lack ofinformation on outreach clinics in general prac-tice and their predicted growth. However, thepilot study was fairly substantial in size andinvolved the collection of data from nine out-reach clinics with outpatient controls acrossEngland. The study included 146 outreachpatients and 148 outpatients. This providessufficient statistical power for comparisons be-

60


tween sites (outreach and outpatients clinics),although statistical power will be weaker forbetween specialty comparisons. The largerscale evaluation of outreach clinics across Eng-land which has recently been launched by theinvestigators will be able to test the differencesreported here between specialities and in alarger sample of clinics.

The authors are grateful to Margaret Hall and Gerald Pope forcoding and data entry, Lesley Marriott for administration, OrlaMurphy, Alison Abery and Marie McClay for assistance withfollowing up late respondents, to the patients, doctors andmanagers who so willingly participated in the study and gavetheir time, to the members of the study's Advisory Group andto our collaborators from the Universities of Manchester andYork: Professor David Wilkin, Mrs Mary Black, Dr BrendaLeese and Mr Toby Gosden. The study was funded by theNHS Management Executive at North Thames Regional HealthAuthority, R&D Division: NHS R&D Programme on the Prim-ary and Secondary Care Interface. Crown Copyright Reserved,1995. The views reported do not necessarily represent those ofthe funding body.

1 National Audit Office. Out-patient serices in the NHS. Lon-don: HMSO, 1991.

2 Hull FM, Westerman RF. Referral to medical out-patientsdepartments at teaching hospitals in Birmingham andAmsterdam. BMJ 1986;293:311-4.

3 Cartwright A, Windsor J. Outpatients and their doctors. Astudy of patients, potential patients, general practitioners andhospital doctors. London: HMSO, 1993.

4 Sullivan FM, Hoare T, Gilmour H. Outpatient clinic re-ferrals and their outcome. BrJ Gen Pract 1992;42:111-15.

5 Dornan C, Fowler G, Mann JI, Markus A, Thorogood MA.A community study of diabetes in Oxfordshire. R CoilGen Pract 1983;33:151-5.

6 Dowie R. General practitioners and consultants. London: KingEdward's Hospital Fund for London, 1983.

7 Grace JF, Armstrong D. Reasons for referral to hospital:extent of agreement between the perceptions of patients,general practitioners and consultants. Family Practice1986;3: 143-7.

8 King's Fund Commission on the Future of London's AcuteHealth Services. London health care 2010: changing the.future ofservices in the capital. London: Kings Fund Centre,1992.

9 Moss F, McNicol M. Secondary care beyond Tomlinson:an opportunity to be seized or squandered. BMJ 1992;305:1211-4.

10 Tomlinson B. Report ofan inquiry into London's health service,medical education and research. London: HMSO, 1992.

11 Department of Health. Making London better. Manchester:Health Publications Unit, 1993.

12 Beardshaw V, Gordon P, Pampling D. Primary care de-velopment zones. BMJ 1993;306:323-5.

13 Bailey JJ, Black ME, Wilkin D. Specialist outreach clinicsin general practice. BMJ 1994;308:1083-6.

14 Harris A. Specialist outreach clinics. More questions thananswers until they have been properly evaluated. Editorial.BMJ 1994;308:1053.

15 Strathdee G, Williams P. A survey ofpsychiatrists in primarycare, the silent growth of a new service. J R Coll Gen Pract1984;34:615-8.

16 Goldberg D, Jackson G. Interface between primary careand specialist mental health care. Editorial. BrJ Gen Pract1992;42:267-9.

17 Creed F, Marks B. Liaison psychiatry in general practice: acomparison of the liaison attachment and shifted out-patient clinic models. J7 R Coll Gen Pract 1989;39:514-7.

18 Davies AR, Ware JE. GHAA' consumer satisfaction surveyand manual. Washington: Group Health Association ofAmerica, 1991.

19 Scott B, Brook RH, Lohr KN, Goldberg GA. Con-ceptualisation and measurement ofphysiologic healthfor adults.Vol 10. Joint disorders. R-2262/10-HHS. Santa Monica,California: The RAND Corporation, 1981.

20 Radosevich DM, Husnik MJ. An abbreviated health statusquestionnaire: the HSQ-12. Update. Bloomington, MN:Newsletter of the Health Outcomes Institute 1995;2:1-4.

21 Bowling A. What things are important in people's lives? Asurvey ofthe public's judgements to inform scales ofhealthrelated quality of life. Soc Sci Med 1995;41: 1447-62.

22 Parkerson GR, Broadhead WE, Tse C-KJ. The Duke se-verity of illness checklist (DUSOI) for measurement ofseverity and comorbidity. J Clin Epidemiol 1993;46:379-93.

23 Bowling A, Jacobson B, Southgate L, Formby J. Generalpractitioners' views of quality specifications for "out-patient referrals and care contracts". BMJ' 1991;303:292-4.

61

Community Health Evaluation clinics general practice in ...eprints.kingston.ac.uk/17255/1/Bowling-A-17255.pdf · Evaluation ofspecialists' outreach clinics in generalpractice year

Documents